The issue was caused because of a faulty change to the auth service during a refactor. This was not expected to have an impact on the runner allocation flow. This was reverted to fix the error. We are also going to add tests to ensure issues of this nature do not creep through.